Esquema de autenticación de contenido de imágenes basado en marcas de agua semi-frágil

This paper presents a block-wise semi-fragile watermarking scheme for image content authentication, with tamper area localization and recovery capability. Before the watermark sequence is generated, the original image is divided into two regions: The Region of Interest (ROI) and the Region of Embedding (ROE), where the ROI is selected manually by the owner of the image and the ROE is rest of the whole image after subtracting region belonging to the ROI. The binary watermark of each ROI block is generated by the DC and the 6 lowest AC components in the zig-zag order of the DCT coefficients of these blocks. These extracted bit sequence are embedded into the LSB of the DCT coefficients of the corresponding ROE blocks which are indicated by a mapping list previously generated using a secret key. In the authentication process, the watermark sequence is extracted from the ROI blocks. Then it is compared with the sequence extracted from the corresponding ROE blocks to determine if the ROI block has been altered; in this case the tampered ROI blocks are recovered replacing them by the extracted watermark sequence from ROE. Simulation results show that the reconstructed image quality and the watermark robustness to JPEG compression are better than the previously reported methods by other authors under the same conditions.